CN100588145C - Transmission method and transmitter - Google Patents

Abstract

A transmission method and a transmitter comprising an antenna system (314, 316) for achieving two transmit paths, a first modulator (304) for modulating a signal to be transmitted into complex modulation symbols, and a second modulator (306) for constructing layered channel symbols as linear combinations of the complex modulation symbols. The second modulator (306) constructs channel symbols by using at least a first and a second non-zero coefficient when performing the linear combination, wherein the ratio of the first and second coefficient is not a real number. The channel symbols are transmitted by using a first non-zero total power for transmission on a first transmit path, and a second non-zero total power for transmission on a second transmit path, wherein the first and second totalpowers are not equal.

Description

Sending method and transmitter

Technical field

The present invention relates to transmitter and sending method in the radio system.The invention particularly relates to the sending method of in transmitter, using a more than transfer path.

Background technology

In telecommunication technology, the transfer path that becomes known for repeating signal causes the interference to the signal that sends.The appearance of this situation and the physical form of transfer path are irrelevant, and no matter it is for example radio link, optical fiber or copper cable.Especially in radio communication, the factor of quality that situation appears at transfer path usually is the different time according to link, and appear at the link duration of existence.

Decline is the typical phenomenon on the radio path, and it causes the change of transmission channel.Other while link also can cause disturbing, and interference can be by the function of time and position.

The solution of this problem is to use diversity in transmitter.Time diversity is used alternation sum coding, the time diversity in the signal that its realization will send.Yet the shortcoming here is in transmission, particularly when channel declines at a slow speed, produces delay.In frequency diversity, signal is sent under some frequencies simultaneously.Yet when channel had wide coherence bandwidth, this was a kind of inefficient methodology.

Antenna diversity uses a more than antenna in signal sends and/or receives.So the signal component by the different channels multipath transmisstion unlikely is subjected to the interference that declines simultaneously.In receive diversity, two or more antennas with diverse location or polarization (polarization) are used to receive the signal that is sent.The shortcoming of receive diversity is to be difficult to be implemented in use two antennas in the small-sized terminals.Concentrate at the transmission branch, same signal uses two or more different antennae to be sent to receiver.Since with provide terminal to compare to be easier to provide the base station with more than antenna, so send diversity than the easier down link that is applicable in the mobile telephone system of receive diversity.

For using a plurality of antennas in the base station and using the situation of an antenna at mobile radio station, up link can be known as SIMO (the many outputs of single input) wireless channel, and down link can be known as MISO (the single output of many inputs) channel.In MIMO (multiple-input and multiple-output) system, on transmitter and receiver, all use a plurality of antennas.In fading channel, this produces transmitter and receiver diversity respectively.MIMO can realize superperformance, but it depends on by the different antennae transmission and receives the signal that reaches by the different channels propagation.In other words, channel will can not have too many relevant each other.

The target of telecommunication technology is not only to send signal as far as possible error-free, and transmits information as far as possible efficiently.Here, efficient is meant that target is to use the capacity of transmission channel in transfer of data as far as possible efficiently.The transmission rate that when the planning honeycomb wireless system, will realize especially noticeable (interesting).

Usually, the raising of the use of diversity and transmission rate is the mutual exclusion candidate.

At H.EI Gamal and M.Damen:An algebraic number the oreticframework for space-time coding, Proc.ISIT, in June, 2002, p.132 in, the precoding of space-time code matrix has been proposed, wherein the Hadamard conversion symbol that is used to rotate.This causes maximum symbol diversity (MSD) pattern.This means that the applying unit matrix sends each single symbol.The prior art precoder is the maximum symbol diversity of the reality of similar above-mentioned Hadamard precoder (MSD), or real and non-MSD.

Summary of the invention

The purpose of this invention is to provide and be used for by using the equipment that diversity realizes the method for high transfer rate and good resistance perturbed force and is used to realize this method that sends.

Based on one aspect of the invention, a kind of sending method is provided, comprise by the linear combination of complex modulation symbol and construct the layering channel symbol, by at least two transfer path transmitting channel symbols, when the structure channel symbol, in at least one layer, when carrying out linear combination, use at least the first and second nonzero coefficients, wherein the ratio of first and second coefficients is not real number, and at least one modulation symbol, uses the first non-zero gross power to send on first transfer path, and use the second non-zero gross power on second transfer path, to send, wherein first and second gross powers are unequal.

Based on another aspect of the present invention, a kind of transmitter is provided, comprise the antenna assembly that is used to realize to be used to transmit two transfer paths of signal, be used for the signal that will send is modulated into the device of complex modulation symbol, be used for constructing the device of layering channel symbol by the linear combination of complex modulation symbol.Transmitter also comprises and being used for by using at least the first and second nonzero coefficients to construct the device of channel symbol at least one layer when carrying out linear combination, wherein the ratio of first and second coefficients is not real number, and be used for by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power on second transfer path, to send at least one modulation symbol, thereby the device of transmitting channel symbol, wherein first and second gross powers are unequal.

Based on another aspect of the present invention, a kind of transmitter is provided, comprise the antenna system that is used to realize to be used for transmit two transfer paths of signal, be used for the signal that will send is modulated into first modulator of complex modulation symbol, be used for constructing second modulator of layering channel symbol by the linear combination of complex modulation symbol.Second modulator is configured to by use the first and second nonzero coefficients at least one layer when carrying out linear combination at least, construct the layering channel symbol, wherein the ratio of first and second coefficients is not real number, and second modulator and antenna system are configured to by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power to send on second transfer path at least one modulation symbol, thereby the transmitting channel symbol, wherein first and second gross powers are unequal.

Based on another aspect of the present invention, a kind of base station transmitter of cellular wireless system is provided, comprise the antenna system that is used to realize to be used for transmit two transfer paths of signal, be used for the signal that will send is modulated into first modulator of complex modulation symbol, be used for constructing second modulator of layering channel symbol by the linear combination of complex modulation symbol.Second modulator is configured to by using the first and second nonzero coefficients to construct the layering channel symbol at least one layer when carrying out linear combination at least, wherein the ratio of first and second coefficients is not real number, and second modulator and antenna system are configured to by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power to send on second transfer path at least one modulation symbol, thereby the transmitting channel symbol, wherein first and second gross powers are unequal.

Based on another aspect of the present invention, a kind of terminal equipment of cellular wireless system is provided, comprise the antenna system that is used to realize to be used for transmit two transfer paths of signal, be used for the signal that will send is modulated into first modulator of complex modulation symbol, be used for constructing second modulator of layering channel symbol by the linear combination of complex modulation symbol.Second modulator is configured to by using the first and second nonzero coefficients to construct the layering channel symbol at least one layer when carrying out linear combination at least, wherein the ratio of first and second coefficients is not real number, and second modulator and antenna system are configured to by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power to send on second transfer path at least one modulation symbol, thereby the transmitting channel symbol, wherein first and second gross powers are unequal.

The present invention relates to a kind of method and apparatus that makes that wireless messages can transmit between discrete unit, two spaces, wherein transmitter has at least two transfer paths at least.Transfer path can use for example omnidirectional antenna or beam-forming any means, realizes under the situation of utilizing or do not utilize channel information.

Solution of the present invention provides some advantages.Some performance measurement that relates to the linear matrix modulator approach is diversity order and coding gain.Diversity order can with the rank correlation of code word matrix of differences:

D ^(ce)＝X ^(c)-X ^(e)， (1)

D wherein ^(ce)Be the code word matrix of differences, it is described and sends code word X ^(c)With reception code word X ^(e)Between difference.Coding gain is considered the absolute value of the determinant (product of nonzero eigenvalue) of code word matrix of differences.For unlimited SNR, usually the performance criteria that uses during idle bit when design is

$\underset{x^{\left(c\right)}\≠x^{\left(e\right)}}{\min }\det D^{\left(\mathrm{ce}\right)H}{D}^{\left(\mathrm{ce}\right)}---\left(2\right)$

Embodiments of the invention keep diversity order same as the prior art, but have higher coding gain.

Use prior art full rate (wherein character rate the equals number of transmit antennas) design of multiple pre-coding matrix to make them need the precoding of homogeneous (homogeneous) symbol, wherein each symbol is rotated so that the power of each symbol is evenly distributed on the transmission resource of all selections.Use pre-coding matrix to produce the MSD precoding, wherein each of matrix (non-zero) element has equal-wattage.This has the shortcoming of raising peak-to-average power than (PAR), its complicated Designing power amplifier.In addition, always do not find optimum diversity or coding gain.Use the solution that is provided to avoid this two kinds of shortcomings.

In the solution of preferred embodiment, in transmitter, reduce PAR, in rayleigh fading channel, improve performance simultaneously to surpass the prior art design.

In an embodiment, realize above purpose by using at least one multiple unit pre-coding matrix, described multiple unit pre-coding matrix comprises at least two nonzero elements with different transmitted powers.These are used to construct the alphabetic(al) layer of high character rate (each channel uses a more than symbol) modulation, and one of them coding dimension is the space, and at least one other dimension obtains from for example time, frequency, code, waveform or Doppler territory.

In an embodiment, be reduced to equidistant QAM (quadrature amplitude modulation) planisphere from different antennae with at the channel symbol that different time sends, so that each antenna just uses a planisphere.Particularly 16-QAM for example by hard-wired situation under, simplified realization.

Description of drawings

Below, with reference to preferred embodiments and drawings the present invention is described in more detail, wherein

Fig. 1 illustrates the structure as the telecommunication system of example,

Fig. 2 illustrates the structure as the telecommunication system of example in more detail,

Fig. 3 illustrates the example of transmitter among the embodiment,

Fig. 4 illustrates the performance of two embodiment of the present invention, and

Fig. 5 A and 5B illustrate the channel symbol planisphere corresponding to two embodiment of the present invention.

Embodiment

Describe the structure of UMTS mobile telephone system with reference to figure 1, this system is as the example that wherein can use the system of the embodiment of the invention.

The major part of mobile telephone system is core network CN, UMTS grounding wireless access network network UTRAN and user equipment (UE).

Interface between core network CN and the Radio Access Network UTRAN is known as Iu, and the air interface between UTRAN and the UE is known as Uu.

User equipment (UE) is made up of two parts: mobile device ME, comprise the wireless terminal that is used on interface Uu, setting up Radio Link, with UMTS subscriber identity module USIM, it is to comprise user ID data and carry out recognizer usually, the smart card of storage encryption parameter and user data.

UTRAN is made up of RNS RNS.RNS is made up of radio network controller (RNC) and one or more Node B.In fact, Node B is meant the base station.Radio network controller (RNC) is by connected BTS management Radio Resource.

Core network CN is made up of several portions.Attaching position register HLR is the database that is used for the user attaching system of maintenance customer's service profile.Attaching position register is also safeguarded the data of relevant customer location with the accuracy of MSC.With regard to circuit switching (CS) service, moving and serving switching center/visitor location register MSC/VLR is switch (MSC) and database (VLR), and it serves terminal equipment.MSC switched circuit Exchange Service, and VLR safeguards the data of relevant user profile and position.Gateway MSC GMSC connects the switch of UMTS to external service or network.All circuit switching connections exchange by GMSC.The function of service GPRS (GPRS) support node SGSN is corresponding to the function of MSC/VLR, but it transmits packet switching (PS) connection.Correspondingly, the function of Gateway GPRS Support Node GGSN is corresponding to GMSC, but this is with regard to packet switching connects.External network can be divided into two kinds: the packet switching network of the circuit-switched network of existing telephone network and for example internet for example.

UTMS comprises some specified interfaces.The Cu interface is between smart card USIM and mobile device ME.The Uu interface is between terminal equipment and base station.Interface between core network CN and the Radio Access Network UTRAN is known as Iu.Interface between the RNS RNS is known as Iur.Soft handover between the radio network controller of this permission different vendor.Interface between radio network controller (RNC) and the base station B is known as Iub.

Fig. 1 quite briefly shows this structure, will be in Fig. 2 to this in more detail the example by cellular wireless system carry out diagram.Fig. 2 only illustrates necessary module, but to it will be understood by those skilled in the art that conventional cellular radio also comprises some other functions and structure, it needn't here be described in more detail.Shall also be noted that Fig. 2 only illustrates exemplary configurations.Details based on system of the present invention can be different from shown in Fig. 2, but this species diversity is nonsensical to the present invention.

Cellular radio generally include fixed network infrastructure (being network portion 200) and can fix, vehicle-mounted or portable terminal equipment 202.Network portion 200 comprises base station 204.The base station is corresponding to the Node B shown in the last accompanying drawing.Several base stations 204 is controlled with centralized system by the radio network controller 206 that communicates with.Base station 204 comprises transceiver 208 and multiplexer module 212.

Base station 204 also comprises control unit 210, the operation of its control transceiver (combination transmitter and receiver) 208 and multiplexer 212.Multiplexer 212 is used to the business and the control channel that are used by some transceivers 208 are placed on the common transfer link 214.Transmission link 214 forms Iub interface.

The transceiver 208 of base station 204 is communicated by letter with antenna element 218, and its double-direction radio that is implemented to terminal equipment 202 connects 216.The structure that will connect 216 frames that send by double-direction radio is at the special appointment of system, and it is known as air interface Uu.In a preferred embodiment of the invention, at least a portion signal passes through three or more transmitting antennas, or is sent by the three or more beams that some transmitting antennas provide.

Radio network controller 206 comprises group switching domain 220 and control unit 222.Group switching domain 220 is used to switched voice and data, and connects signaling circuit.The RNS 232 that is formed by base station 204 and radio network controller 206 also comprises code converter 224.Because voice can thereby transmit with the cellular network form,, thereby save transmission capacity so code converter 224 is usually located at the approaching as far as possible position of serving switching center 228 of moving between code converter 224 and radio network controller 206.

The different digital speech coding form that uses between code converter 224 conversion public switched telephone networks and the mobile telephone network is with compatible each other, other form of some from the fixed network format conversion to cellular radio network format for example, and vice versa.Control unit 222 is carried out and is called out control, mobile management, statistic gathering and signaling.

As shown in Figure 2, group switching domain 220 is used to serve that switching center 228 and gateway MSC 230 carry out the exchange of public switched telephone network (PSTN) 236 and to the exchange of packet transmission networks 242 by moving.

Connection between packet transmission networks 242 and the group switching domain 220 is set up by SGSN (Serving GPRS Support Node) 240.The function of support node 240 is to transmit grouping between base station system and GGSN (Gateway GPRS Support Node) 244, and keeps the record of the position of terminal equipment 202 in its zone.

Gateway node 244 connects public packet transmission network 246 and packet transmission networks 242.Internet protocol or X.25 agreement can be used for interface.The internal structure of gateway node 244 encapsulating packets transmission networks 242, so that it is hidden public packet transmission network 246, and therefore, public packet transmission network 246 is regarded packet transmission networks 242 as subnet, and the public packet transmission network can addressed packet to the terminal equipment 202 that is arranged in network, and receive grouping from terminal equipment 202.

Packet transmission networks 242 typically uses Internet protocol and transmits signaling and the private network of the user data that the tunnel transmits.Under Internet protocol layer, the architecture of network configuration 242 and agreement can change based on operator.Public packet transmission network 246 can be a fhe global the Internet for example.

Fig. 3 illustrates the more specific example of transmitter.Transmitter can be the wireless terminal of subscriber equipment for example, or the transmitter of base station transceiver.As known to persons skilled in the art, transmitter also can comprise for example other parts of filter, but for simply, does not describe them here.

Transmitter comprises information source 300, and it is with assigned source speed R _SourceSend information bit b.These positions are admitted to encoder 302, and this encoder also can be carried out staggered.Usually, encoder comprises having code rate R _cBinary coder.The output of coding comprises bits of coded, and it is further sent into modulator 304.Modulator is modulated into the multiple modulation vector x to the bits of coded of specified quantity.

The output of modulator 304 also can be known as modulation symbol.Modulation symbol is the plural number of directly being constructed by the position.For example, the QPSK symbol is to constitute by 2, and the 16-QAM symbol is to constitute by 4.Here, QPSK is meant Quadrature Phase Shift Keying, and 16-QAM is meant 16 quadrature amplitude modulation.These are typical modulator approaches of using in the transmitting system.

Modulator 306 when the output of modulator is connected to sky when operation.Let as assume that institute's modulation signal will be by the transmission resource T and the transmission beam N of basic quadrature _bTransmit.Transmit resource for example can comprise to, subcarrier or extended code.The character rate of modulator is R when empty _sThereby, when sky in the modulator, R _sThe sequence of T symbol is mapped to TxN _bMatrix X.Modulator also can comprise precoding when empty.

For example, use linear MIMO or MISO modulation, modulator is configured to the Linear Mapping modulation symbol to empty time-code matrix X when empty.The length of piece is T when empty.

The output of modulator 306 also can be known as channel symbol when empty.Channel symbol is to be described in the symbol period from the phase place of transfer path (for example antenna or beam) transmission and the plural number of amplitude.

In the code of preferred embodiment, channel symbol uses linear transformation to be made of modulation symbol.

The output signal 308 of modulator thereby comprise N when empty _bIndividual parallel flow, and it is sent to beam formation device 310.Beam forms the appointment N of device 310 by transmitter _tIndividual space is transmitted resource and is set up N _bIndividual transfer path.These resources can be antennas for example, antenna beam, or available polarization method.Beam forms the operation of device can use N _bXN _tThe matrix W diagram.Beam forms the output signal 312 of device thereby comprises N _tIndividual parallel flow, and it is converted into used wireless frequency, and in amplifier 314, amplify, and use antenna system 316 to send.

Transmitter can also comprise the controller 318 of the operation of controlling transmitter.

Be used for from the gross power that given antenna or antenna beam send a modulation symbol be linear combination coefficient multiply by modulation symbol (being used to construct the channel symbol that sends from antenna or antenna beam) square summation.

In a preferred embodiment, use the hierarchical matrix modulation.Layer is the set by the channel symbol of one group of modulation symbol structure.These modulation symbols are not used when any other layer of structure.If in layer, there is a more than channel symbol, then for each channel symbol of constructing by first set of modulation symbol, in this layer, there is at least one other channel symbol, it makes first and second set have at least one modulation symbol jointly by the second set structure of modulation symbol.

Suppose an embodiment, wherein transmitter has two transfer paths, and receiver has at least two antennas (or antenna beam) and speed is 2.Thereby matrix has dimension 2 * 2 when empty.Let us supposes that also pre-coding matrix is 2 * 2 unit matrixs

$U(\mathrm{\μ},\mathrm{\ν})=\left[\begin{array}{cc}\mathrm{\μ}& \mathrm{\υ}\\ -{\mathrm{\υ}}^{*}& {\mathrm{\μ}}^{*}\end{array}\right]---\left(3\right)$

And matrix is used to the symbolic vector of transport layer k ${\tilde{x}}_k=U_{x_k}.$ In this embodiment, use at least two transfer paths (antenna, antenna beam).By sign map is used the layering sign matrix to the transmission matrix that uses a more than symbol layer.At least two symbol operation parameters at least one layer are by precoding, wherein | μ | ≠ | ν |, | μ |, | ν | ≠ 0, and alternatively, at least one number is a plural number (promptly having the non-zero imaginary part).

In addition, different layers can have different pre-coding matrixes, as long as at least one layer is observed above-mentioned constraint.Symbol in the different layers can have different capacity and/or symbol constellation (for example scalar rotation).Empty time-code matrix modulation has following form:

$X=\left[\begin{array}{cc}\mathrm{\μ}{x}_1+\mathrm{\υ}{x}_2& e^{\mathrm{j\φ}+\mathrm{\π}/4}(\mathrm{\μ}{\mathrm{\α}}_3+\mathrm{\υ}{x}_4)\\ e^{\mathrm{j\φ}+\mathrm{\π}/4}({\mathrm{\μ}}^{*}{x}_4-{\mathrm{\υ}}^{*}{x}_3)& {\mathrm{\μ}}^{*}{x}_2-{\mathrm{\υ}}^{*}{x}_1\end{array}\right]---\left(4\right)$

φ+π/4th wherein, the rotation between the layer, | μ | ≠ | ν |, μ and ν's is real number one of at least.The modulator of equation (4) is based on the threading design of Hadamard.

In another embodiment, can use the modulator of i threading, have following form

$X=\left[\begin{array}{cc}\mathrm{\μ}{x}_1+\mathrm{\υ}{x}_2& e^{\mathrm{j\φ}}(\mathrm{\μ}{x}_3+\mathrm{\υ}{x}_4)\\ {\mathrm{je}}^{\mathrm{j\φ}}({\mathrm{\μ}}^{*}{x}_4-{\mathrm{\υ}}^{*}{x}_3)& {\mathrm{\μ}}^{*}{x}_2-{\mathrm{\υ}}^{*}{x}_1\end{array}\right]---\left(5\right)$

Wherein | μ | ≠ | ν |.

The performance of (4) and (5) is identical in rayleigh fading channel, can cut apart between upper and lower off-diagonal bight by using simple unit's similarity conversion at the j in the lower left corner of (5), and this does not change performance.The result has the form that presents in (4).The advantage of i threading design is that optimum non-MSD precoding has φ=0 usually.This allows us all the time two antennas to be used the channel symbol of identity set.

In the modulator of equation (4) and (5), there are two layers.By modulation symbol x ₁And x ₂Linear combination structure to angleplied laminate, and by modulation symbol x ₃And x ₄Construct non-to angleplied laminate.

More than two matrixes have non-MSD form.When use | μ | ≠ | ν | non-MSD pre-coding matrix the time, not only PAR reduces (comparing with the MSD design), and has improved performance in addition.Compare with the pattern based on Clifford of prior art, the shortcoming of the matrix that is provided is to have increased detection complexity slightly, particularly under the situation of using accurate quadrature aided detection method.On the other hand, non-MSD coding method has improved the performance that detects based on the maximum likelihood (ML) of reducing the number of the tree reduction.This is drawn by non-MSD characteristic, wherein realizes that for allocated channel some symbol is encoded better than other symbol.The reduction searching algorithm of ML can directly use this feature.

In one embodiment of the invention, the i of modulation matrix X exercises to use by beam and forms vectorial w _iThe given beam that characterizes sends.Here, w _iBe the M dimensional vector, wherein M is the quantity of the antenna element of use in transmitting.One forms vectorial w _iIt can be characteristic vector corresponding to the eigenvalue of maximum of channel correlation matrix.Respectively, another vectorial w ₂Can be characteristic vector corresponding to second eigenvalue of maximum, or the like.In addition, transmit matrix column and can form vector correlations with different beams, promptly in above-mentioned example, the strongest beam w ₁Can be assigned to some other row that transmits matrix.The characteristic that transmits matrix is depended in this association usually.

In one embodiment of the invention, the symbol at the different coordinates place in all layers of vector x can be selected from different modulating alphabet or its rotation.For example, the BPSK of 90 degree rotations modulates on the Q axle, and conventional BPSK modulates on the I axle.For example, the part symbol can come from the QPSK alphabet, and part comes from the BPSK alphabet.Thereby, can use different bit rate.For example, under the QPSK-BPSK situation, throughput can be on the ground floor 2 and on the second layer 1, and total bit rate will be 3bps/Hz.

In one embodiment of the invention, form one group of modulation matrix based on above-described principle.For example, sign condition Ux ₁(first pair of angleplied laminate) and Ux ₂(the second opposition angleplied laminate) is configured, and one group of transmission matrix uses the QPSK symbol constellation to be formed by these sign conditions.Thereby we obtain 2 ⁴* 2 ⁴=2 ⁸Individual modulation matrix, each needs two symbol periods so that each symbol period sends log ₂(2 ⁸)/2=4 position.In other words, 8 positions are used to select in many modulation matrixs, and each specific 8 bit sequence is corresponding to the mark (selection) relevant with mapping.This mark can be determined by the mark of the stream in two layers fully.Yet in an embodiment, other labeling method also can be used.After forming matrix, each transmits matrix can be by mark again.This again mark can be used as the expansion of the matrix of Gray mark, for example guaranteeing that approximating different to transmit matrixes relevant with similar mark by euclidian metric, and vice versa.Obviously, other labeling method can be used, and does not have the restriction to input alphabet, even above-mentioned example uses the QPSK incoming symbol.Notice that also two identical or different input planispheres can have identical or different mark, thereby produce the mark of the transmission matrix of matrix value.

The performance of comparing two embodiment of the solution that is proposed with the prior art solution is provided at Fig. 4.Fig. 4 contrasts Eb/No simulation BER is shown.This accompanying drawing is provided at 2 and takes advantage of the analog result that 4bps/Hz transmits in the 2 smooth mimo channels.Prior art solution 400 comes from M.Damen, K.Abed-Meraim and J.-C.Belfiore, Diagonal algebraicspace-time block code, IEEE Trans.Inform.Theory, vol.48, no.3, pp.628-636, in March, 2002.First embodiment 402 expands 16-QAM, and second embodiment 404 is plural non-MSD.People such as Damen disclose the MSD notion, and the solution that is proposed has non-MSD type.Suppose and use two transmissions and two reception antennas.The parameter of three notions of comparing is in form 1.Parameter comprises μ, ν and φ.In addition, form 1 illustrates the performance of PAR (peak averaging value ratio) and the MIN DET that calculates based on equation (2).Fig. 4 shows with MSD 400 and compares that non-MSD notion 402,404 has improved performance really.Equally, we notice that MIN-DET is measuring quite reliably performance.

Notion	Mark	μ	ν	φ	MIN DET	PAR
							People such as Damen	400	1/√2	e jπ/4/√2	π/8	0.17	1.71
Expand 16-QAM	402	0.8507	0.526	0	0.89	1.89
							The non-MSD of plural number	404	0.8881	e jπ/40.4597	π/8	0.62	1.58

Form 1

Describe in Fig. 5 A corresponding to the channel symbol planisphere that expands 16-QAM situation 402, wherein use the modulator of equation (4).Channel symbol is to expand the 16-QAM symbol really.At two transmitting antennas and two signaling stages planisphere is described.Channel symbol planisphere corresponding to the non-MSD situation 404 of plural number is described respectively in Fig. 5 B.

If selecting the precoding parameter is μ=2/ √ 5 and ν=1/ √ 5, then MIN-DET and performance are traded off slightly, but the channel symbol planisphere is fully equidistant 16-QAM.Be apparent that use the modulator of equation (5) and mentioned parameter, each antenna sends 16-QAM all the time; Do not need rotation, this realization brings the loss of utmost point low performance.

Suppose an embodiment, wherein transmitter has four transfer paths, and receiver has four antennas (or antenna beam), and speed is 2.Starting point is the pattern based on Clifford of following form:

$X_{\mathrm{IDABBA}}=\frac{1}{\sqrt{2}}\left[\begin{array}{cc}{X}_A+X_C& X_B+X_D\\ j(X_B-X_D)& X_A-X_C\end{array}\right]---\left(6\right)$

X wherein _A, X _B, X _CAnd X _DHave the STTD form (or have corresponding to any quadrature the form of space-time block code (square)):

$X_A({\mathrm{<figure-callout\; id="1"\; label="x"\; filenames="C2004800339730021H2.png,C2004800339730022H2.png"\; state="\{\{state\}\}">x</figure-callout>}}_1,x_2)=\left[\begin{array}{cc}{x}_1& x_2\\ -x_2^{*}& x_1^{*}\end{array}\right]$ $X_B(x_3,x_4)=\left[\begin{array}{cc}{x}_3& x_4\\ -x_4^{*}& x_3^{*}\end{array}\right]$

$X_C(x_5,x_3)=\left[\begin{array}{cc}{x}_5& x_6\\ -x_6^{*}& x_5^{*}\end{array}\right]$ $X_D(x_7,x_8)=\left[\begin{array}{cc}{x}_7& x_8\\ -x_8^{*}& x_7^{*}\end{array}\right]$

Use rotation STTD piece X _CAnd X _DThe simple precoder of middle information symbol can make equation (6) full rank:

X _{3，4，7，8}→e ^jφX _{3，4，7，8}

In an embodiment, the non-MSD form of modulation matrix can be formulated as:

$X_{\mathrm{IDABBA}\ln \mathrm{on}}=\left[\begin{array}{cc}\cos \mathrm{\&rho;}{X}_A+\sin \mathrm{\&rho;}{X}_C& \cos \mathrm{\&rho;}{X}_B+\sin \mathrm{\&rho;}{X}_D\\ j(\sin \mathrm{\&rho;}{X}_B-\cos \mathrm{\&rho;}{X}_D)& \sin \mathrm{\&rho;}{X}_A-\cos \mathrm{\&rho;}{X}_C\end{array}\right]---\left(7\right)$

As symbolically, equation (7) becomes

$X_{\mathrm{IDABBA}\ln \mathrm{on}}=\left[\begin{array}{cccc}\cos \mathrm{\&rho;}{x}_1& \cos \mathrm{\&rho;}{x}_2& \cos \mathrm{\&rho;}{x}_3& \cos \mathrm{\&rho;}{x}_4\\ -\cos \mathrm{\&rho;}{x}_2^{*}& \cos \mathrm{\&rho;}{x}_1^{*}& -\cos \mathrm{\&rho;}{x}_4^{*}& \cos \mathrm{\&rho;}{x}_3^{*}\\ j\sin \mathrm{\&rho;}{x}_3& j\sin \mathrm{\&rho;}{x}_4& \sin \mathrm{\&rho;}{x}_1& \sin \mathrm{\&rho;}{x}_2\\ -\sin \mathrm{\&rho;}{x}_4^{*}& \sin \mathrm{\&rho;}{x}_3^{*}& -\sin \mathrm{\&rho;}{x}_2^{*}& \sin \mathrm{\&rho;}{x}_1^{*}\end{array}\right]$

$+\left[\begin{array}{cccc}\sin \mathrm{\&rho;}{x}_5& \sin \mathrm{\&rho;}{x}_6& \sin \mathrm{\&rho;}{x}_7& \sin \mathrm{\&rho;}{x}_8\\ -\sin \mathrm{\&rho;}{x}_6^{*}& \sin \mathrm{\&rho;}{x}_5^{*}& -\sin \mathrm{\&rho;}{x}_8^{*}& \sin \mathrm{\&rho;}{x}_7^{*}\\ -j\cos \mathrm{\&rho;}{x}_7& -j\cos \mathrm{\&rho;}{x}_8& -\cos \mathrm{\&rho;}{x}_5& -\cos \mathrm{\&rho;}{x}_6\\ j\cos \mathrm{\&rho;}{x}_8^{*}& -j\cos \mathrm{\&rho;}{x}_7^{*}& \cos \mathrm{\&rho;}{x}_6^{*}& \cos \mathrm{\&rho;}{x}_5^{*}\end{array}\right]---\left(8\right)$

In the modulator of equation (8), there are four layers.To angleplied laminate by modulation symbol x ₁And x ₅, i.e. cornerwise element cos ρ x of first matrix ₁, cos ρ x ₁ ^*, sin ρ x ₁, sin ρ x ₁ ^*With the element sin ρ x on the diagonal of second matrix ₅, sin ρ x ₅ ^*,-cos ρ x ₅, cos ρ x ₅ ^*Linear combination structure.

A (1,2), (2,1), (3,4) (4,3) layer is by modulation symbol x ₂And x ₆Linear combination structure.Here (x y) is illustrated in x row, the matrix element on y is capable.Thereby for example (1,2) represents first entry of a matrix element-cos ρ x ₂ ^*, and second entry of a matrix element-sin ρ x ₆ ^*A (1,3), (3,1), (2,4) (4,2) layer is by modulation symbol x ₃And x ₇Linear combination structure.At last, oppose that angleplied laminate is by modulation symbol x ₄And x ₈Linear combination structure.

Optimum non-MSD precoder does not need the symbol rotation.Thereby all channel symbols also are to expand the 16-QAM planisphere, and in a preferred embodiment, all channel symbols also are pure equidistant 16-QAM symbols.Have the MSD of different parameters and the characteristic of non-MSD iDABBA and be collected into form 2.Having nearly by maximization, the MIN-DET of the paired error event of 4 faults carries out optimization.Error event with minimum bit mistake has minimum euclid distance, so diversity protection (by the MIN-DET indication) is most important to them.MIN-DET (4) and MIN-DET indicate the minimum determinant of finding of the faults that reaches 4 faults and any number respectively.

Notion	cosρ	φ	MIN DET	MIN DET(4)	PAR
						opt(4)MSD	1	0.36	0.00043	2	1.94
The non-MSD of opt (4)	0.8881	0	0.0032	2.67	1.82
						16-QAM	0.8944	0	N/A	2.56	1.8

Form 2

Suppose an embodiment, wherein transmitter has four transfer paths, and receiver has four antennas (or antenna beam), and speed is 4.This transmission can be constructed by two modulators of parallel transmission form (5):

X _DITHRD＝[X _ITHRD(x ₁，x ₂，x ₃，x ₄)，X _ITHRD(x ₅，x ₆，x ₇，x ₈)](9)

As representing that with symbol equation (9) becomes

$X_{\mathrm{DITHRD}}=\left[\begin{array}{cccc}\mathrm{\&mu;}{x}_1+\mathrm{\&upsi;}{x}_2& e^{\mathrm{j\&phi;}}(\mathrm{\&mu;}{x}_1+\mathrm{\&upsi;}{x}_2)& \mathrm{\&mu;}{x}_5+\mathrm{\&upsi;}{x}_6& e^{\mathrm{j\&phi;}}(\mathrm{\&mu;}{x}_7+\mathrm{\&upsi;}{x}_8)\\ {\mathrm{je}}^{\mathrm{j\&phi;}}({\mathrm{\&mu;}}^{*}{x}_4+{\mathrm{\&upsi;}}^{*}{x}_3)& {\mathrm{\&mu;}}^{*}{x}_2-{\mathrm{\&upsi;}}^{*}{x}_1& {\mathrm{je}}^{\mathrm{j\&phi;}}({\mathrm{\&mu;}}^{*}{x}_8-{\mathrm{\&upsi;}}^{*}{x}_7)& {\mathrm{\&mu;}}^{*}{x}_6-{\mathrm{\&upsi;}}^{*}{x}_5\end{array}\right]---\left(10\right)$

In the modulator of equation (10), there are four layers similarly.(1,1), (2,2) layer is by modulation symbol x ₁And x ₂Linear combination structure, (1,2), (2,1) layer is by modulation symbol x ₃And x ₄Linear combination structure, and (1,3), (2,4) layer is by modulation symbol x ₅And x ₆Composite construction.(1,4), (2,3) layer is by modulation symbol x ₇And x ₈Composite construction.

In an embodiment, the non-MSD iDABBA modulation matrix by equation (7) reduces (punctured) two transmitting antennas.Resulting transmission is equivalent to two antenna beam transmission of equation (5).By series connection from antenna to two of 1-2 and 3-4 such transmission, can construct speed 4 transmission.Thereby by reducing (puncturing) antenna 2 and 4 by (7), matrix of consequence is

$X_{\mathrm{IDABBApuno}}=\left[\begin{array}{cc}\cos \mathrm{\&rho;}{x}_1+\sin \mathrm{\&rho;}{x}_5& \cos \mathrm{\&rho;}{x}_3+\sin \mathrm{\&rho;}{x}_7\\ \cos \mathrm{\&rho;}{x}_2+\sin \mathrm{\&rho;}{x}_6& \cos \mathrm{\&rho;}{x}_4+\sin \mathrm{\&rho;}{x}_8\\ j(\sin \mathrm{\&rho;}{x}_3-\cos \mathrm{\&rho;}{x}_7)& \sin \mathrm{\&rho;}{x}_1-\cos \mathrm{\&rho;}{x}_5\\ j(\sin \mathrm{\&rho;}{x}_4-\cos \mathrm{\&rho;}{x}_8)& \sin \mathrm{\&rho;}{x}_2-\cos \mathrm{\&rho;}{x}_6\end{array}\right]---\left(11\right)$

Signal can use signal model Y=XH+ noise to simulate, and wherein H is a matrix of describing transmission channel, and Y is the received signal matrix.Thereby the reduction antenna represents to reduce the row of X.Signal model hereto, true STTD can be expressed as

$\left[\begin{array}{cc}{x}_1& -x_2^{*}\\ x_2& x_1^{*}\end{array}\right]$

Row 1 and 3 is non-MSD modulators of equation (5) with odd-numbered symbol, and row 2 and 4 provides the identical modulator with even-numbered symbol.From performance perspective, if ρ=arccos (2/ √ 5), it provides equidistant 16-QAM channel symbol, and equation (5) then, the modulator of (7) and (9) are almost optimum.

In an embodiment, channel symbol is configured to the linear combination of complex modulation symbol in the modulator when sky.These channel symbols are sent out by at least two transfer paths.At at least one modulation symbol, the first gross power non-zero that is used on first transfer path, sending, and the second gross power non-zero that is used on second transfer path, sending, and first and second gross powers are unequal.

Above-mentioned feature has realized the non-MSD characteristic of modulation matrix.For example, suppose that two antennas or antenna beam use in conjunction with the modulator of equation (4) and (5), are used to send modulation symbol x ₁Gross power from antenna 1 | μ | ²With from antenna 2 | ν | ²In non-MSD solution, | μ | ²≠ | ν | ²

When using the modulator of equation (8), be used to send modulation symbol x ₁Gross power from antenna 1 and 2 (cos ρ) ²With from (the sin ρ) of antenna 3 and 4 ²In non-MSD solution, (cos ρ) ²≠ (sin ρ) ²

When using the modulator of equation (10), be used to send modulation symbol x ₁Gross power from antenna 1 | μ | ², from antenna 2 | ν | ²With from 0 of antenna 3 and 4.In these embodiment of non-MSD solution, zero is omitted.Need | μ | ²≠ | ν | ²

In an embodiment, when passing through the linear combination structure channel symbol of complex modulation symbol in the modulator when at sky, at least one layer, at least the first and second nonzero coefficients are used to linear combination.In an embodiment, the ratio of first and second coefficients is not real number.

Above-mentioned feature realizes multiple precoding characteristic.For example, in the modulator of realizing based on equation (4), ratio μ/ν, μ/μ ^*, ν/ν ^*Or ν/μ ^*In at least one be not real number.This means that at least one parameter is a plural number.Should be noted that when taking the ratio of non-coefficient to angleplied laminate layer multiplication coefficient e ^{J φ+π/4}Be eliminated.

In two hierarchical modulator of equation (5), be μ/ν to the ratio of the coefficient of angleplied laminate, μ/μ ^*, ν/ν ^*And ν/μ ^*, but not be μ/ν to the ratio of the coefficient of angleplied laminate, j μ/μ ^*, j ν/ν ^*With j ν/μ ^*In this case, when calculating non-ratio to angleplied laminate, layer multiplication coefficient e ^{J φ}Be eliminated, but a j keeps.Thereby in an embodiment, any one is non-real number to angleplied laminate ratio, and at least one that means μ or ν is plural number, and/or non-be non-real number to one of angleplied laminate ratio, this is to any μ, ν all sets up, no matter whether they are real numbers.

Similarly, in the modulator of equation (8), in (1,3), (3,1), and in the layer of (2,4) (4,2), ratio cos ρ/jcos ρ occurs, and it is non-real number to any real number ρ.Modulator for equation (9) also is like this.

In another embodiment, pre-coding matrix is a real matrix, but makes the power difference in the layer internal symbol that different time sends be at least 20% pair 80%.

Even reference has been described the present invention in the above based on the example of accompanying drawing, yet obviously the present invention is not limited, and can revise in a number of ways within the scope of the appended claims.

Claims (12)

1, a kind of data transmission method for uplink comprises:

As the linear combination of modulation symbol construct the layering channel symbol and

By at least two transfer path transmitting channel symbols,

It is characterized in that

When the structure channel symbol, at least one layer, when carrying out linear combination, use at least the first and second nonzero coefficients, wherein the ratio of first and second nonzero coefficients is not real number, and

At at least one modulation symbol, use the first non-zero gross power on first transfer path, to send, and use the second non-zero gross power on second transfer path, to send, wherein the first and second non-zero gross powers are unequal.

2, data transmission method for uplink as claimed in claim 1 is characterized in that using at least one the multiple pre-encoder matrix that comprises described first and second nonzero coefficients, and wherein said first and second nonzero coefficients have different transmitted powers.

3, data transmission method for uplink as claimed in claim 1 is characterized in that using at least one real pre-encoder matrix, and wherein the transmitted power difference between the modulation symbol of different time is at least 20% pair 80% in layer.

4, data transmission method for uplink as claimed in claim 1 is characterized in that the channel symbol that uses different transfer paths and different time to send forms equidistant qam constellation figure.

5, data transmission method for uplink as claimed in claim 1 is characterized in that the channel symbol that uses different transfer paths and different time to send forms grid.

6, data transmission method for uplink as claimed in claim 5 is characterized in that grid is equidistant.

7, a kind of transmitter comprises:

Be used to realize to be used to transmit the antenna assembly (314,316) of two transfer paths of signal,

Be used for the signal that will send be modulated into modulation symbol device (304) and

Be used for by when carrying out linear combination, using at least the first and second nonzero coefficients, construct the device (306) of layering channel symbol as the linear combination of modulation symbol at least one layer, wherein the ratio of first and second nonzero coefficients be not real number and

Be used for by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power on second transfer path, to send at least one modulation symbol, thereby the device (306 of transmitting channel symbol, 310,314,316), wherein the first and second non-zero gross powers are unequal.

8, a kind of transmitter comprises:

Be used to realize to be used for transmit the antenna system (314,316) of two transfer paths of signal,

Be used for the signal that will send is modulated into first modulator (304) of modulation symbol,

Be used for constructing second modulator (306) of layering channel symbol as the linear combination of modulation symbol,

It is characterized in that

Second modulator (306) is configured to construct channel symbol by use at least the first and second nonzero coefficients at least one layer when carrying out linear combination, and wherein the ratio of first and second nonzero coefficients is not real number, and

Second modulator (306) and antenna system (310,314,316) be configured to by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power on second transfer path, to send at least one modulation symbol, thereby the transmitting channel symbol, wherein the first and second non-zero gross powers are unequal.

9, as claim 7 or 8 described transmitters, it is characterized in that transmitter comprises to be used for by using described first and second nonzero coefficients to come the device of transmitting channel symbol, wherein said first and second nonzero coefficients have different transmitted powers.

10, as claim 7 or 8 described transmitters, it is characterized in that transmitter comprises is used for by using at least one real pre-encoder matrix to come the device of transmitting channel symbol, and the transmitted power difference between the modulation symbol of different time is at least 20% pair 80% in layer in this matrix.

11, a kind of base station transmitter of cellular wireless system comprises:

Be used to realize transmit in week the antenna system (314,316) of two transfer paths of signal,

Be used for the signal that will send is modulated into first modulator (304) of modulation symbol,

Be used for constructing second modulator (306) of layering channel symbol as the linear combination of modulation symbol,

It is characterized in that

Second modulator (306) is configured to by using the first and second nonzero coefficients to construct channel symbol at least one layer when carrying out linear combination at least, and wherein the ratio of first and second nonzero coefficients is not real number, and

Second modulator (306) and antenna system (310,314,316) be configured to by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power on second transfer path, to send at least one modulation symbol, thereby the transmitting channel symbol, wherein the first and second non-zero gross powers are unequal.

12, a kind of terminal equipment of cellular wireless system comprises:

Be used to realize to be used for transmit the antenna system (314,316) of two transfer paths of signal,

Be used for the signal that will send is modulated into first modulator (304) of modulation symbol,

Be used for constructing second modulator (306) of layering channel symbol as the linear combination of modulation symbol,

It is characterized in that

Second modulator (306) is configured to by using the first and second nonzero coefficients to construct channel symbol at least one layer when carrying out linear combination at least, and wherein the ratio of first and second nonzero coefficients is not real number, and

Second modulator (306) and antenna system (310,314,316) be configured to by use the first non-zero gross power sending on first transfer path and using the second non-zero gross power on second transfer path, to send at least one modulation symbol, thereby the transmitting channel symbol, wherein the first and second non-zero gross powers are unequal.

Images